Oxidative Bioactivation of Crotyl Alcohol to the Toxic Endogenous

Jun 25, 2002 - To facilitate study of the mechanisms underlying CA toxicity, we ... Moreover, in mouse hepatocytes, crotyl alcohol produced marked tim...
0 downloads 0 Views 182KB Size
Chem. Res. Toxicol. 2002, 15, 1051-1058

1051

Oxidative Bioactivation of Crotyl Alcohol to the Toxic Endogenous Aldehyde Crotonaldehyde: Association of Protein Carbonylation with Toxicity in Mouse Hepatocytes Frank R. Fontaine,† Rachael A. Dunlop,† Dennis R. Petersen,‡ and Philip C. Burcham*,† Molecular Toxicology Research Group, Department of Clinical & Experimental Pharmacology, Adelaide University, Adelaide, SA 5005, Australia, and Department of Pharmaceutical Sciences, University of Colorado Health Sciences Center, Denver, Colorado Received January 17, 2002

Recent confirmation that the toxic unsaturated aldehyde crotonaldehyde (CA) contributes to protein damage during lipid peroxidation confers interest on the molecular actions of this substance. However, since a plethora of structurally related aldehydes form during membrane oxidation, clarifying the toxicological significance of individual products (e.g., CA) is challenging. To facilitate study of the mechanisms underlying CA toxicity, we explored the possibility that it can be formed enzymatically from an unsaturated precursor, crotyl alcohol. This is analogous to the way allyl alcohol is converted in vivo to its toxic oxidation product, acrolein. In kinetic studies, we found that crotyl alcohol was readily oxidized by equine liver alcohol dehydrogenase, with electrospray-mass spectrometry confirming that CA was the main product formed. Moreover, in mouse hepatocytes, crotyl alcohol produced marked time- and concentrationdependent cell killing as well as pronounced glutathione depletion. Both cytotoxicity and glutathione loss were abolished by the alcohol dehydrogenase inhibitor 4-methylpyrazole, indicating an oxidation product mediated these effects. In keeping with expectations that carbonyl-retaining Michael addition adducts would feature prominently during protein modification by CA, exposure to crotyl alcohol resulted in marked carbonylation of a wide range of cell proteins, an effect that was also abolished by 4-methylpyrazole. Damage to a subset of small proteins (e.g., 29, 32, 33 kDa) closely correlated with the severity of cell death. Collectively, these results demonstrate that crotyl alcohol is a useful tool for studying the biochemical and molecular events accompanying intracellular CA formation.

Introduction Like its 3-carbon counterpart acrolein, the 4-carbon R,β-unsaturated aldehyde crotonaldehyde (CA)1 is a common environmental pollutant, formed upon combustion of fossil fuels and wood (1, 2). Inhalation of tobbacco smoke is also a significant source of CA, exposing smokers to 30-170 µg/kg/day (3). CA also occurs naturally in certain fruits and vegetables, selected dairy products, and some alcoholic beverages (1, 4). In addition to these exogenous sources, CA is also formed endogenously during lipid peroxidation (LPO), a factor that may contribute to the aldehydes’ presence in the serum of nonsmokers (5, 6). As with other unsaturated aldehydes, CA exhibits many toxicological properties, including an ability to * Correspondence should be addressed to this author at the Molecular Toxicology Research Group, Department of Clinical & Experimental Pharmacology, Adelaide University, Adelaide, SA 5005, Australia. Phone: 61-8-8303-5287, Fax: 61-8-8224-0687, Email: [email protected]. † Adelaide University. ‡ University of Colorado Health Sciences Center. 1 Abbreviations: ADH, alcohol dehydrogenase; CA, crotonaldehyde; CrOH, crotyl alcohol; 2,4-DNPH, 2,4-dinitrophenylhydrazine; 2,4-DNP, 2,4-dinitrophenyl; LDH, lactate dehydrogenase; LPO, lipid peroxidation; 4-MP, 4-methylpyrazole; MS, mass spectrometry.

cause multi-organ toxicity in laboratory animals; DNA damage and mutagenicity in vitro; and carcinogenicity in vivo (2-4, 7-9). As a reactive bifunctional electrophile, CA readily generates adducts in a range of cell macromolecules. For example, cyclic CA-derived adducts with deoxyguanosine have been detected in the genome of several species including humans (7). In recent work by Uchida and associates, an immunohistochemical approach identified CA-adducted proteins in the kidneys of prooxidant-treated rats, although on account of the cross-reactivity of the antibodies used, the possibility that higher-order 2-alkenals (e.g., 2-pentenal) might have contributed to the damage could not be excluded (6). These workers also characterized several adducts formed during modification of proteins by CA (6). As with other unsaturated aldehydes, the chemistry of protein modification by CA was diverse, involving carbonyl-retaining β-substituted butanal adducts at lysine and histidine residues, as well as a cyclized adduct that formed via sequential Michael additions to the -amino group of lysine, N-(2,5-dimethyl-3-formyl-3,4-dehydropiperidino)lysine (6). Finally, Uchida et al. also identified a novel pyridinium adduct [N-(5-ethyl-2-methylpyridinium)lysine] as the major antigenic adduct formed by CA. This adduct also formed via addition of two CA molecules to

10.1021/tx0255119 CCC: $22.00 © 2002 American Chemical Society Published on Web 06/25/2002

1052

Chem. Res. Toxicol., Vol. 15, No. 8, 2002

a single lysine group, although in this case a Schiff-type reaction occurred after the initial Michael addition (6). Such findings raise questions as to the toxicological significance of protein damage by CA, and whether it is associated with the pathogenesis of CA-mediated cell death. To begin studying these issues in an isolated cell model, we explored a method for the in situ generation of CA directly within the subcellular environment, using an unsaturated alcohol (crotyl alcohol, CrOH) as a metabolic precursor (10). This approach is analogous to the use of allyl alcohol to examine mechanisms of cell injury by acrolein, which forms via alcohol dehydrogenase (ADH)-dependent oxidative metabolism (11). Similarly, Tappel and associates demonstrated that 1,3-propanediol is a useful in vivo precursor to another toxic lipid-derived aldehyde, malondialdehyde (12). More recently, Neely et al. reported that an enzyme-activated triester analogue facilitated study of the cellular effects of another important lipid-derived unsaturated aldehyde, 4-hydroxy-2nonenal (13). In an in vitro setting, such approaches offer the advantage of minimizing “quenching” of aldehydes by nucleophilic culture media constituents (e.g., amino acids, glutathione, etc.) that occur when aldehydes are directly added to the media (13). In this study, we show that equine liver ADH oxidizes CrOH to CA with comparable efficiency to that with which it converts allyl alcohol to acrolein. We also explored CrOH metabolism in isolated mouse hepatocytes, and found it undergoes ADH-catalyzed conversion to CA. Formation of CA was accompanied by marked glutathione depletion, protein carbonylation, and cell death. We conclude that CrOH is a useful tool for studying the molecular events underlying CA-mediated cell injury.

Materials and Methods Animals. Male albino Swiss mice (30-35 g) were obtained from the Adelaide University Animal Breeding Facility (Waite Institute, SA, Australia). All animals had ad libitum access to water and standard rodent food and were housed at 22 °C on a 12 h light/dark cycle. Chemicals and Reagents. Collagenase (Clostridium histolycum, Type IV), 4-methyl pyrazole, RPMI-1640 medium, o-phthaldialdehyde (OPT), bovine serum albumin (BSA, fraction V), 2,4-dinitrophenylhydrazine (DNPH), and equine liver alcohol dehydrogenase (ADH, EC 1.1.1.1) were purchased from Sigma Chemical Co. (St. Louis, MO). Fatty acid contaminants of BSA were removed using activated charcoal prior to its inclusion in culture media (14). Crotyl alcohol (1:19 cis:trans, 97% purity) was purchased from Fluka Chemika, Germany. NAD+ was purchased from Boehringer Mannheim, Germany. All other reagents were of analytical grade. Kinetics of ADH-Catalyzed CrOH Oxidation. The ADHcatalyzed oxidation of CrOH was followed by measuring NADH generation at 340 nm, using a UV-spectrophotometric method adapted to suit a microplate reader (Biolumin 960, Molecular Dynamics, CA) (15). Fresh solutions of NAD+, ADH, ethanol, allyl alcohol, and CrOH were prepared each day in Tris (0.2 M)-KCl (40 mM) buffer (pH 7.3). The concentration range (1, 2, 2.5, 3, 5, and 10 mM) of each alcohol studied was selected on the basis of results from preliminary experiments in which they yielded reproducible apparent initial velocity data (Km and Vmax) upon construction of Lineweaver-Burk plots. To each well of a 96-well plate were added ADH (final concentration 0.15 unit/ mL) and the various substrate alcohols to give a final volume of 150 µL. Reactions were started by the addition of 100 µL of NAD+ solution, to give a final concentration of 1 mM. All reactions were performed in triplicate. After mixing, reactions were allowed to proceed for 30 min at 37 °C, with absorbance

Fontaine et al. measurements taken every 2 min. Data were collected using “Xperiment Software” (Molecular Dynamics, Sunnyvale, CA). Estimates of kinetic parameters were obtained via Microsoft Excel. Averages calculated from the initial linear region of ∆A/ min curves (from triplicate determinations) were used to calculate 1/V. Lineweaver-Burk plots were constructed using Microsoft Excel. In one experiment, to definitively confirm that CA was produced during CrOH oxidation, aliquots of enzymatic reaction mixture were analyzed via electrospray-MS using a Finnigan LCQ ion trap mass spectrometer. Mouse Hepatocyte Isolation and Incubations. Hepatocytes were prepared by collagenase digestion of mouse liver (16). The viability of the cells was assessed via trypan blue exclusion, with typically 75-85% of the cells excluding the dye. The cells were washed 3 times in CaCl2-supplemented Krebs Henseleit. After the final wash, cells were resuspended in RPMI-1640 media supplemented with 1% BSA, 0.03% L-glutamine, and penicillin/streptomycin (50 units/mL and 50 µg/mL, respectively) and then layered onto 60 mm collagen-coated culture dishes at a density of 1.5 × 106 cells/dish. Cells were then placed in a humidified 5% CO2 incubator at 37 °C for 2-3 h to allow cell attachment. The monolayers were then washed 3 times with PBS before fresh medium containing CrOH was added. In some instances, cells were preincubated for 30 min with the alcohol dehydrogenase inhibitor 4-methylpyrazole (4-MP, 500 µM) prior to CrOH addition (17). Both CrOH and 4-MP were dissolved directly in culture media, with solutions prepared immediately prior to use. Experiments were then continued for up to 240 min, with aliquots of culture media removed at regular intervals for the determination of LDH activity. For dishes that were used in glutathione or protein carbonylation assays, experiments were concluded after 30 min. HPLC Procedures. To verify CA formation from CrOH during cell incubations, the aldehyde was trapped as a 2,4dinitrophenylhydrazone and then detected via HPLC (18). Briefly, a 75 µL aliquot of cell culture medium was collected from cells that had been exposed to a range of CrOH concentrations for 30 min as described above. Next, an equalivalent volume of DNPH reagent was added to each sample [DNPH reagent was prepared by diluting a 5 mM stock solution of 2,4DNPH (in 2 N HCl) with 4 volumes of the Tris/KCl buffer specified above]. After allowing the derivatization to proceed for 15 min at room temperature, the samples were diluted 1:9 in 70% (v/v) aqueous acetonitrile. Following centrifugation for 5 min at 13000g, a 100 µL sample was analyzed via reverse-phase HPLC using a Beckman Ultrasphere ODS 5 µm column (250 × 4.6 mm) which was eluted at a flow rate of 1 mL/min. The mobile phase comprised 70% (v/v) aqueous acetonitrile. The CA-derived 2,4-DNP-hydrazone was detected at 370 nm using a HewlettPackard 1100 UV detector. A standard curve was prepared by reacting equivalent volumes of various concentrations of authentic CA (2-1000 µM) with 2,4-DNPH. Under these conditions, the retention time for 2,4-DNP-derivatized CA was 6.0 min. To confirm that all the CA produced from CrOH was derivatized by 2,4-DNPH, column eluent from parallel runs was monitored at 210 nm to detect any underivatized CA (retention time approximately 3.0 min). Biochemical Assays. Cellular glutathione levels were determined using the fluorometric method of Hissin and Hilf (19). Briefly, monolayers were washed 3 times with cold PBS before 1.5 mL of 3% perchloric acid was added to each dish. The dish contents were then transferred to an Eppendorf tube before they were centrifuged for 5 min at 3000g. The glutathione content of 100 µL aliquots of supernatant was then determined using a standard curve made from a standard solution of glutathione prepared in 3% perchloric acid. Cellular glutathione levels were expressed as micrograms of glutathione per milligram of protein. To standardize glutathione levels, perchloric acid precipitates were dissolved in 500 µL of 0.5 M NaOH, and then the Pierce Coomassie Plus Protein Assay Kit (Rockford, MA) was used to estimate protein concentrations. LDH activity in culture media was measured using a spectrophotometric method (20). LDH

Crotyl Alcohol Bioactivation and Protein Carbonylation

Chem. Res. Toxicol., Vol. 15, No. 8, 2002 1053

Table 1. Km and Vmax Values for the Oxidation of Ethanol, Allyl Alcohol, and Crotyl Alcohol by Equine Liver Alcohol Dehydrogenasea species

substrate

Km (µM)

Vmax [nmol h-1 (mg of protein)-1]

r2 value

horse

ethanol allyl alcohol crotyl alcohol

2460 390 710

780 190 140

0.9313 0.9938 0.9587

a Correlation coefficients (r2) for the initial velocity estimates for each alcohol are also displayed. Km, Vmax values, and correlation coefficients were obtained from 3 independent observations.

activities at each time point were expressed as a percentage of the total activity in individual dishes, determined at the end of experiments after lysing cells with Triton X-100. Methods for the immunodetection of protein carbonyls were as we described recently (11). Data Analysis. Data were analyzed via repeat measures ANOVA using GraphPad Prism (v 3.02, GraphPad Software Inc., San Diego, CA). Tukey’s post hoc test was used to detect differences between groups. A significance level of p < 0.05 was used.

Results To investigate the suitability of CrOH as a metabolic precursor for CA, we first determined whether it is a substrate for alcohol dehydrogenase (ADH). Using commercially available enzyme (equine liver) in a 96-well assay format, we compared ADH’s ability to metabolize CrOH to the readiness with which it oxidizes two established ADH substrates, allyl alcohol and ethanol. Substrate oxidation was assessed by following NADH formation at 340 nm. The kinetics of substrate oxidation by ADH typically conform to the Ordered Bi-Bi model, which (sequentially) involves substrate binding followed by NAD binding, product release, and finally release of NADH from the enzyme complex (21-23). When depicted in double-reciprocal form, kinetic data from such reactions exhibit a linear relationship between initial reaction velocity versus substrate concentration (i.e., LineweaverBurk plot) (21). Analysis of our kinetic data for CrOH, allyl alcohol, and ethanol oxidation by ADH matched these expectations, yielding correlation coefficients ranging from 0.93 to 0.99 for the three substrates when plotted as a double-reciprocal Lineweaver-Burk plot (Table 1). The Km and Vmax estimates for each of the three substrates indicate that the efficiency of ADH-catalyzed oxidation of CrOH was comparable to that of allyl alcohol (Table 1). To obtain definitive proof for CA formation from CrOH, we subjected an aliquot of enzymatic reaction mixture (after a 60 min reaction with 0.1 mM CrOH) to analysis via electrospray-MS. In keeping with expectations, the mass spectrum was dominated by an ion at m/z 69, corresponding to the deprotonated molecular anion of CA (data not shown). When this ion was further fragmented via MS/MS, a daughter ion at m/z 40 was generated, corresponding to cleavage R to the carbonyl group with concomitant loss of CHO•. The intensity of the latter ion was directly dependent on collisional energy used during MS/MS (data not shown). Since the spectra obtained upon MS/MS analysis of enzyme-generated versus synthetic CA were superimposable, we concluded that CA is the major product of enzymatic CrOH oxidation. To assess the efficiency of CA formation from CrOH in a simple cell-free model, we used 2,4-dinitrophenylhydrazine (DNPH) to trap the aldehyde in enzyme

Figure 1. Extracellular formation of CA in mouse hepatocytes exposed to various concentrations of CrOH (100-500 µM) for 30 min. An aliquot of culture medium (75 µL) was mixed with an equivalent volume of DNPH reagent (1 mM), and then HPLC analysis was performed as outlined under Materials and Methods. The retention time for the CA-derived 2,4-dinitrophenylhydrazone was 6.0 min (cf. arrows). For the sake of clarity, the portion of the profile corresponding to the elution time for the CA-DNP-hydrazone is shown for each CrOH concentration, while the inset depicts a representative full chromatogram.

reaction mixtures, followed by reverse-phase HPLC to detect the resulting CA-derived 2,4-DNP-hydrazone. CA yields were estimated via a standard curve prepared by carrying authentic CA (2-1000 µM) through the same assay steps. The efficiency of recovery of CA-derived 2,4DNP-hydrazone from incubation mixtures “spiked” with 5 nmol of CA was at least 99% (data not shown). After a 60 min incubation of 0.1 mM CrOH with equine liver ADH, the yield of CA was 16 ( 1.1%, while at 1 and 10 mM CrOH, concentrations above the Km for the enzyme, conversion rates were 3.4 ( 0.44% and 0.55 ( 0.06%, respectively (N ) 3, means ( SE). Upon extended incubation, CA formation appeared to reach a plateau after 1-2 h. Consequently, after 6 h, the conversion efficiencies were just 18 ( 3% (0.1 mM CrOH), 6.4 ( 0.71% (1 mM CrOH), and 1.1 ( 0.12% (10 mM CrOH) (n ) 3 independent experiments). Such observations are typical of Ordered Bi-Bi kinetics, since reaction products can act as uncompetitive inhibitors of the enzyme complex (21). Also, rate-limiting kinetics for ADH-catalyzed reactions may reflect the formation of abortive enzymeNADH-alcohol complexes (22, 23). To determine whether CA is formed from CrOH in intact cells, we used the above 2,4-DNPH-based HPLC method to detect CA in aliquots of culture media from mouse hepatocytes following a 30 min incubation with a range of CrOH concentrations (Figure 1). Under the HPLC conditions described under Materials and Methods, the resulting CA-DNP-hydrazone eluted with a retention time of 6.0 min (Figure 1). The UV-HPLC profiles shown in Figure 1 confirm that CrOH is a true metabolic precursor for CA, since a clear concentrationdependent increase in CA levels was seen in the culture media of liver cells exposed to 100-500 µM CrOH for 30

1054

Chem. Res. Toxicol., Vol. 15, No. 8, 2002

Fontaine et al.

Figure 2. (Panel A) Time course for lactate dehydrogenase (LDH) leakage from mouse hepatocyte monolayers exposed to 0 (‚‚‚0‚‚‚), 100 (/), 200 (4), 300 (]), 400 (O), or 500 µM (- ‚‚0- ‚‚) CrOH. Aliquots of media were taken at different times and assayed for LDH activity. (Panel B) Hepatocyte GSH contents after 30 min treatment with various concentrations of CrOH. In both panels, each data point represents the mean ( SEM of 3 independent observations. A “*” designates p < 0.05, “**” p < 0.01, and “***” p < 0.001 when compared to control.

min (Figure 1). In keeping with the likelihood that the CrOH concentration range studied was below the Km for the murine alcohol dehydrogenase involved in CrOH oxidation, the yield of extracellular CA was essentially the same at 0.1 mM CrOH (25.4%) as at 0.5 mM CrOH (26.4%). The finding of these high concentrations of CA in the culture medium of CrOH-treated cells indicates that a significant proportion of CA derived from CrOH is sufficiently stable to exit the cell in which it is formed. The data in Figure 1 further suggests that CA is the main aldehyde formed upon CrOH biotransformation. In early work, Comporti and associates identified 4-hydroxynonenal and several other saturated and unsaturated aldehydes in liver extracts from mice poisoned with allyl alcohol, the metabolic precursor of acrolein (24). This raised the possibility that a range of biogenic aldehydes might form during exposure of hepatocytes to CrOH. However, the HPLC profiles in Figure 1 indicate that none of the corresponding DNP-hydrazones generated by a range of lipid-derived aldehydes (namely, malondialdehyde, acrolein, 4-hydroxynonenal, trans-2-pentenal, and trans-2-hexenal) were present in culture media of CrOH-treated cells (the retention times for the DNPhydrazones formed from the respective aldehydes were 4.4, 4.8, 6.4, 7.4, and 9.8 min). Given these findings, it is safe to conclude that the major cellular effects of CrOH are mediated by CA rather than other reactive unsaturated aldehydes. To examine the toxicological consequences of CA formation from CrOH, the time course for the leakage of lactate dehydrogenase (LDH) into culture media was determined in mouse hepatocytes incubated with 100500 µM concentrations of CrOH for up to 4 h (Figure 2A). LDH leakage is a sensitive indicator of cytotoxicity in rodent hepatocytes (25). CrOH was very toxic to the cells, producing marked time- and concentration-dependent increases in LDH leakage (Figure 2A). For example, 400 and 500 µM CrOH produced 65 and 82% LDH leakage within just 180 min (Figure 2A). For all CrOH concentrations examined, no cytotoxicity was evident at the first time point at which LDH activity was assessed (i.e., 30 min, Figure 2A). Nevertheless, even within this brief

exposure period, CrOH caused significant glutathione depletion, with 300 µM and higher concentrations of CrOH decreasing glutathione levels to approximately 40% of controls (Figure 2B). Whether CrOH toxicity in mouse hepatocytes was mediated by the oxidation product CA was then examined using the alcohol dehydrogenase inhibitor 4-MP to block any role of this pathway in CrOH biotransformation. The results shown in Figure 3A indicate that 4-MP strongly suppressed the cytotoxicity of 400 µM CrOH in mouse hepatocytes, diminishing LDH leakage to levels comparable to those in controls (Figure 3A). Moreover, 4-MP completely prevented the glutathione depletion associated with a 30 min exposure to 400 µM CrOH (Figure 3B). By establishing that glutathione depletion required the oxidative metabolism of CrOH, this rules out an involvement of the R,β-unsaturated bond of CrOH in conjugative reactions with glutathione. In allyl alcohol-treated mouse hepatocytes, the formation of acrolein caused pronounced protein carbonylation prior to the loss of cell integrity (11). This reflects the readiness with which short-chain 2-alkenals such as acrolein modify proteins, forming Michael-addition adducts that retain an intact carbonyl group. To see whether protein carbonylation also occurs in the early stages of CrOH toxicity, cells were exposed to 100-500 µM concentrations of CrOH for 30 min after which protein extracts were analyzed for carbonylation status using an immunochemical assay based on adduct derivatization with 2,4-DNPH (26, 27). The data in Figure 4A indicate extensive protein carbonylation occurred during this brief exposure period. In general, the degree of protein carbonylation produced by a given concentration of CrOH correlated closely with the degree of cytotoxicity produced by that concentration of the alcohol (Figure 2A). Thus, the nominally toxic concentrations of 100 and 200 µM CrOH produced only a modest increase in the carbonyl content of cell proteins, while the three higher, more toxic concentrations produced pronounced carbonylation of a wide range of proteins (Figure 4A). Interestingly, the pattern of protein carbonylation produced by the most toxic concentrations of CrOH (300 µM

Crotyl Alcohol Bioactivation and Protein Carbonylation

Chem. Res. Toxicol., Vol. 15, No. 8, 2002 1055

Figure 3. (Panel A) Effect of a 30 min pretreatment with the alcohol dehydrogenase inhibitor 4-MP (0.5 mM) on LDH leakage from mouse hepatocytes exposed to 400 µM CrOH. Cells were preincubated with 4-MP (0.5 mM) for 30 min prior to the addition of CrOH. The various treatments were (0) 400 µM CrOH, (4) CrOH + 500 µM 4-MP, and (]) 4-MP only. Each data point represents the mean ( SEM of 3 independent observations. A “***” indicates p < 0.001 compared to the effect of 400 µM CrOH alone. (Panel B) Effect of 4-MP pretreatment on CrOH-induced GSH depletion in mouse hepatocytes. Cells were exposed to 400 µM CrOH alone or in the presence of 0.5 mM 4-MP for 30 min prior to the determination of GSH. Each point represents the mean ( SEM of 3 independent observations. A “*” indicates p < 0.05 when compared to control, while “∝” designates p < 0.05 when compared to 400 µM CrOH.

Figure 4. (Panel A) Western blot showing carbonylation of mouse hepatocyte proteins (25 µg/lane) following a 30 min exposure to various concentrations of CrOH. Lane 1, control cells; lane 2, 100 µM CrOH; lane 3, 200 µM CrOH; lane 4, 300 µM CrOH; lane 5, 400 µM CrOH; lane 6, 500 µM CrOH. Hepatocyte proteins (25 µg/lane) were derivatized with 2,4-DNPH and resolved on a 4-20% polyacrylamide gel as described under Materials and Methods. (Panel B) Results of the densitometric analysis of the carbonyl content of three selected proteins (29, 32, and 33 kDa) showing strong concentration-dependent increases in immunostaining relative to controls.

and higher) suggests a subset of small-to-medium-sized proteins is particularly prone to carbonylation (Figure 4A, compare lane 1 vs lanes 4-6). For example, densitometric analysis of three selected bands (29, 32, and 33 kDa protein) revealed that 500 µM CrOH increased the immunostaining of these bands some 3- 5-fold over and above the damage produced by 100 µM CrOH (Figure 4B). A central role for CA in CrOH toxicity was confirmed during study of the effect of 4-MP on protein carbonylation in CrOH-exposed hepatocytes (Figure 5). As in the preceding experiment, the period of exposure to CrOH was just 30 min. Once again, 300 µM CrOH strongly increased the carbonyl content of a broad range of proteins (Figure 5, lane 3 vs lane 1). However, in cells that were pretreated with 4-MP, 300 µM CrOH did not produce its characteristically strong increase in protein carbonylation (Figure 5, lane 4). Together with the finding that 4-MP also prevented glutathione depletion and LDH leakage in CrOH-treated cells, this suggests protein carbonylation is a direct consequence of CA formation from CrOH.

Discussion Since a wide range of potentially toxic aldehydes form during LPO, attributing precise cellular events to any particular product is challenging. In the current work, we have shown that the early cellular effects of one reactive LPO-derived aldehyde, CA, can be studied using an unsaturated alcoholic precursor, CrOH. The latter underwent rapid ADH-catalyzed conversion within the intracellular environment to an intermediate that caused glutathione depletion, protein modification, and cytotoxicity. Since this approach enabled study of the cellular effects of CA without the need to add the aldehyde to the culture media of cells, it seems to more closely approximate the endogenous situation whereby CA forms directly within the cellular environment as a result of oxidative damage to cell membranes (6). Given that CA and acrolein are closely related in chemical terms, unsaturated alcohol precursors provide a novel means of comparing the toxic potency of these aldehydes. Allyl alcohol is a classic hepatotoxicant widely used for studying the biochemical mechanisms involved in liver damage. The liver injury produced by allyl alcohol

1056

Chem. Res. Toxicol., Vol. 15, No. 8, 2002

Figure 5. Western blot showing the effect of 4-MP (0.5 mM) pretreatment on carbonylation of mouse hepatocyte proteins following a 30 min exposure to 300 µM CrOH. The treatments in the respective lanes were as follows: lane 1, control cells; lane 2, 0.5 mM 4-MP pretreatment only; lane 3, 300 µM CrOH only; lane 4, 300 µM CrOH + 0.5 mM MP. Hepatocyte proteins (25 µg/lane) were derivatized with 2,4-DNPH and resolved on a 4-20% polyacrylamide gel as described under Materials and Methods.

reflects the readiness with which it undergoes rapid ADH-oxidized conversion to acrolein, particularly within periportal regions in rodent liver (15, 28, 29). We recently characterized the contribution of protein carbonylation to the toxicity of allyl alcohol in isolated mouse hepatocytes, and established that extensive protein damage preceded the irreversible loss of cell integrity (11). Comparison of the results from that study to those obtained for CrOH in the present work reveals clear differences between the toxic potencies of allyl alcohol versus CrOH in isolated mouse liver cells. Thus, the concentration of CrOH that produced 50% LDH leakage after a 3 h incubation period in this study (approximately 350 µM) was around 7-fold higher than the allyl alcohol concentration that produced the same degree of toxicity in the earlier work (approximately 50 µM) (11). Although one possible explanation for the differing toxic potencies of the two alcohols could be that CrOH is less efficiently oxidized than its shorter chain counterpart, the finding in the present study that kinetic parameters (Km and Vmax) for the oxidation of AA and CrOH by equine liver ADH were of the same order of magnitude suggests the efficiency of 2-alkenal formation from comparable concentrations of each alcohol would be similar. Consequently, the differences between the toxic potency of allyl alcohol and CrOH most likely reflect the fact that acrolein is inherently more reactive (electrophilic) than CA. In this regard, in comparisons of the carbonylating potencies of acrolein and CA toward a model protein (bovine serum albumin, 2 h modifications at neutral pH), we have found that the yield of carbonyl adducts produced by acrolein is typically 3-4-times greater than that produced by the equivalent concentration of CA (F. R. Fontaine, unpublished observation). The ADH-dependent metabolic activation of CrOH resulted in pronounced carbonylation of a wide range of cell proteins, an effect that clearly preceded the loss of cell viability (i.e., protein modification was extensive

Fontaine et al.

within just 30 min of commencing exposure to CrOH, Figure 4A). The finding that several small proteins (e.g., 29, 32, and 33 kDa) appeared to be major targets for carbonylation may be significant given that Hartley and associates also observed that a group of comparably sized proteins were targeted by malondialdehyde during carbon tetrachloride toxicity in rat hepatocytes (30). One explanation for these findings could be that these proteins are rich in nucleophilic residues (e.g., cysteine, lysine) and thus might function as cytoprotective “aldehyde sequestering proteins” during oxidative stress (30). Distinguishing between this scenario and the alternative possibility that damage to such proteins actively contributes to disease pathogenesis is a challenging question that no doubt awaits identification of these proteins. The demonstration that CA can be generated intracellularly in a time- and concentration-dependent manner from CrOH opens up new approaches to identifying critical subcellular targets for CA. For example, preparation of radiolabeled CrOH might facilitate detection of adducted proteins in treated animals, which if followed by MS-based approaches would allow rapid identification of adducted proteins (31). Such an approach should reveal whether enzymes that are known to be sensitive to inhibition by CA in in vitro systems (e.g., aldehyde dehydrogenase, cytochrome c reductase) are also damaged by this toxic aldehyde in vivo (32, 33). Moreover, it may reveal whether the loss of CYP450 activity seen previously in the livers of crotonaldehyde-treated male F-344 rats is due to selective modification of these microsomal proteins (34). In a similar vein, since it permits control of the amount of CA generated within the cellular environment, CrOH may prove an invaluable tool for investigating the fate of CA-adducted proteins in cellular systems [e.g., investigation of the role of proteasomal pathways in protein degradation (35)]. CrOH may also facilitate study of the “downstream” biochemical events that occur after CA formation and protein adduction. For example, it could be valuable for determining whether CA elicits the same spectrum of molecular effects that have been identified for acrolein [e.g., inhibition of activation of transcription factors such as nuclear factor κB (NF-κB) and activator protein 1 (AP1) (36)]. Similarly, CrOH may be useful for establishing whether CA induces dysregulation of the expression of genes of known toxicological significance. For example, in recent work using DNA microarray technology, Ulrich and associates found that allyl alcohol produced marked changes in gene expression profiles in the livers of exposed rats (37). Intriguingly, by comparing changes in gene expression profiles in the livers of rats that received toxic doses of 15 diverse hepatotoxicants, the compound with which allyl alcohol clustered most closely was the centrilobular hepatotoxicant carbon tetrachloride (37). The genes disrupted by both compounds included several that participate in cellular responses to oxidative stress, such as heme oxidase, p38-activated mitogen kinase, cytochrome b558, and glutathione S-transferase (37). Since oxidative membrane damage is conspicuous in carbon tetrachloride hepatotoxicity, it is tempting to speculate that aldehydic LPO products might trigger some of the changes in gene expression seen, thus explaining close clustering of gene responses in allyl alcohol- and carbon tetrachloride-treated animals (37). A tool such as CrOH might prove highly useful in addressing whether other unsaturated aldehydes (i.e., CA in this case) induce

Crotyl Alcohol Bioactivation and Protein Carbonylation

comparable changes in gene expression in vivo. Given that CrOH exhibits pronounced ADH-dependent toxicity in freshly isolated mouse hepatocytes, it is safe to predict that this compound will prove hepatotoxic in intact rodents. However, before CrOH could be used in future in vivo studies, it is important that the in vivo hepatotoxic potential of CrOH is first defined in rats and mice. Such information is currently lacking, although a single early study did report findings regarding CrOH biotransformation in rats (38). The major urinary metabolite in CrOH-treated animals was 3-hydroxy-1methylpropylmercapturic acid, with 2-carboxy-1-methylethylmercapturic acid formed in lesser quantities. Importantly, the same metabolites predominated in the urine of rats treated directly with CA itself, confirming that CrOH does indeed undergo efficient conversion to CA in vivo (38). In conclusion, our current work demonstrates that the toxic lipid peroxidation product CA can be generated in an ADH-catalyzed reaction from an unsaturated alcohol precursor, CrOH. In keeping with expectations, CrOH was highly toxic to isolated mouse hepatocytes, producing dose- and time-dependent toxicity that was blocked by prior treatment of cells with an ADH inhibitor, 4-MP. In addition to blocking cytotoxicity, 4-MP also prevented GSH depletion and protein carbonylation in CrOHtreated hepatocytes. These collective findings indicate that CrOH permits CA formation directly within the internal environment of cells, and thus is a useful tool for clarifying the molecular events underlying the toxicity of this aldehyde.

Acknowledgment. This work was partially supported by the National Health & Medical Research Council of Australia (Project Grant 104848, P.C.B. and F.R.F.) and NIH/NIAAA Grant 09300 (D.R.P.). This publication is based on a poster presentation delivered at the 40th Annual Meeting of the Society of Toxicology, San Fransisco, March 2001. The authors are grateful for the assistance of Simon Pyke and Phil Clements during the mass spectrometry experiments reported in this paper.

References (1) IARC (1995) Crotonaldehyde. In IARC Monographs on the evaluation of carcinogenic risks to humans. Dry cleaning, some chlorinated solvents and other industrial chemicals, World Health Organisation and International Agency for Research on Cancer, U.K. (2) Schuler, B. D., and Eder, E. (2000) Development of a 32P-postlabeling method for the detection of 1,N2-propanodeoxyguanosine adducts of crotonaldehyde in vivo. Arch. Toxicol. 74, 404-414. (3) Eder, E., Schuler, D., and Budiawan (1999) Cancer risk assessment for crotonaldehyde and 2-hexenal: an approach. IARC Sci. Publ. 150, 219-232. (4) Budiawan, and Eder, E. (2000) Detection of 1,N2-propanodeoxyguanosine adducts in DNA of Fisher 344 rats by an adapted 32Ppost labeling technique after per os application of crotonaldehyde. Carcinogenesis 21, 1191-1196. (5) Zlatkis, A., Poole, C. F., Brazeli, R., Bafus, D. A., and Spencer, P. S. (1980) Volatile metabolites in sera of normal and diabetic patients. J. Chromatogr. 182, 137-145. (6) Ichihashi, K., Toshihiko, O., Toyokuni, S., and Uchida, K. (2001) Endogenous formation of protein adducts with carcinogenic aldehydes. J. Biol. Chem. 276, 23903-23913. (7) Burcham, P. C. (1998) Genotoxic lipid peroxidation products: their DNA damaging properties and role in formation of endogenous DNA adducts. Mutagenesis 13, 101-119. (8) Neudecker, T., Eder, E., Deininger, C., and Henschler D. (1989) Crotonaldehyde is mutagenic in Salmonella typhimurium TA100. Environ. Mol. Mutagen. 14, 146-148.

Chem. Res. Toxicol., Vol. 15, No. 8, 2002 1057 (9) Marnett, L. J., Hurd, H. K., Hollstein, M. C., Levin, D. E., Esterbauer, H., and Ames, B. N. (1985) Naturally occurring carbonyl coumpounds are mutagens in Salmonella tester strain TA104. Mutat. Res. 148, 25-34. (10) Cederbaum, A. I., Pietrusko, R., Hempal, J., Becker, F. F., and Rubin, E. (1975) Characterization of a nonhepatic alcohol dehydrogenase from rat hepatocellular carcinoma and stomach. Arch. Biochem. Biophys. 171, 348-360. (11) Burcham, P. C., and Fontaine, F. (2001) Extensive protein carbonylation precedes acrolein-mediated cell death in mouse hepatocytes. J. Biochem. Mol. Toxicol. 15, 309-316. (12) Summerfield, F. W., and Tappel, A. L. (1984) Cross-linking of DNA liver and testes of rats fed 1,3-propanediol. Chem.-Biol. Interact. 50, 87-96. (13) Neely, M. D., Amarnath, V., Weitlauf, C., and Montine, T. J. (2002) Synthesis and cellular effects of an intracellularly activated analogue of 4-hydroxynonenal. Chem. Res. Toxicol. 15, 40-47. (14) Chen, R. F. (1967) Removal of fatty acid from serum albumin by charcoal treatment. J. Biol. Chem. 242, 173-181. (15) Rikans, L. E., and Moore, D. R. (1987) Effect of age and sex on allyl alcohol hepatotoxicity in rats: role of liver alcohol and aldehyde dehydrogenase activities. J. Pharmacol. Exp. Ther. 243, 20-26. (16) Harman, A. W., McCamish, L. E., and Henry, C. A. (1987) Isolation of hepatocytes from postnatal mice. J. Pharmacol. Methods 17, 157-63. (17) Hartley, D. P., and Petersen, D. R. (1997) Co-metabolism of ethanol, ethanol-derived acetaldehyde, and 4-hydroxynonenal in isolated rat hepatocytes. Alcohol. Clin. Exp. Res. 21, 298-304. (18) Esterbauer, H., Cheeseman, K. H., Dianzani, M. U., Poli, G., and Slater, T. F. (1982) Separation and characterization of the aldehydic products of lipid peroxidation stimulated by ADP-Fe2+ in rat liver microsomes. Biochem. J. 208, 129-140. (19) Hissin, P. J., and Hilf, R. (1976) A fluorometric method for determination of oxidized and reduced glutathione in tissues. Anal. Biochem. 74, 214-226. (20) Richards, A. H., Lubinski, R. M., and Vanderlinde, R. E. (1975) Studies on the kinetic assay of lactate dehydrogenase activity. Clin. Chem. 21, 1018-1023. (21) Segel, I. H. (1993) Enzyme kinetics: Behavior and analysis of rapid equilibrium and steady-state enzyme systems, J. Wiley & Sons, New York. (22) Pietruszko, R. (1973) Mammalian liver alcohol dehydrogenases. Adv. Exp. Med. Biol. 56, 1-29. (23) Pietruszko, R. (1982) Alcohol dehydrogenase from horse liver, steroid active SS isozyme. Methods Enzymol. 89, 428-434. (24) Comporti, M. (1998) Lipid peroxidation and biogenic aldehydes: from the identification of 4-hydroxynonenal to further achievements in biopathology. Free Radical Res. Commun. 28, 623-635. (25) Moldeus, P., Hogberg J., and Orrenius S. (1978) Isolation and use of liver cells. Methods Enzymol. 52, 60-71. (26) Shacter, E., Williams, J. A., Stadtman, E. R., and Levine, R. L. (1996) Determination of carbonyl groups in oxidised proteins. In Free Radicals, A Practical Approach (Punchard, N. A., and Kelly, F. J. Eds.) pp 159-170, IRL Press, Oxford. (27) Keller, R. J., Halmes, N. C., Hinson, J. A., and Pumford, N. R. (1993) Immunochemical detection of oxidized proteins. Chem. Res. Toxicol. 6, 430-433. (28) Serafini-Cessi, F. (1972) Conversion of allyl alcohol into acrolein by rat liver. Biochem. J. 128, 1103-1107. (29) Smith, P. F., Fisher, R., Shubat, P. J., Gandolfi, A. J., Krumdieck, C. L., and Brendel, K. (1987) In vitro cytotoxicity of allyl alcohol and bromobenzene in a novel organ culture system. Toxicol. Appl. Pharmacol. 87, 509-522. (30) Hartley, D. P., Kroll, D. J., and Petersen, D. R. (1997) Prooxidantinitiated lipid peroxidation in isolated rat hepatocytes: detection of 4-hydroxynonenal- and malondialdehyde-protein adducts. Chem. Res. Toxicol. 10, 895-905. (31) Qiu, Y., Benet, L. Z., and Burlingame, A. L. (1998) Identification of the hepatic protein targets of reactive metabolites of acetaminophen in vivo in mice using two-dimensional gel electrophoresis and mass spectrometry. J. Biol. Chem. 273, 1794017953. (32) Cooper, K. O., Witmer, C. M., and Witz G. (1987) Inhibition of microsomal cytochrome c reductase activity by a series of R,βunsaturated aldehydes. Biochem. Pharmacol. 36, 627-631. (33) Mitchell, D. Y., and Petersen, D. R. (1993) Inhibition of rat liver mitochondrial and cytosolic aldehyde dehydrogenases by crotonaldehyde. Drug Metab. Dispos. 21, 396-399. (34) Cooper, K. O., Witz, G., and Witmer, C. (1992) The effects of R,βunsaturated aldehydes on hepatic thiols and thiol-containing enzymes. Fundam. Appl. Toxicol. 19, 343-349.

1058

Chem. Res. Toxicol., Vol. 15, No. 8, 2002

(35) Merker, K., Sitte, N., and Grune, T. (2000) Hydrogen peroxidemediated protein oxidation in young and old human MRC-5 fibroblasts. Arch. Biochem. Biophys. 375, 50-54. (36) Kehrer, J. P., and Biswal, S. S. (2000) The molecular effects of acrolein. Toxicol. Sci. 57, 6-15. (37) Waring, J. F., Jolly, R. A., Ciurlionis, R., Lum, P. Y., Praestgaard, J. T., Morfitt, D. C., Buratto, B., Roberts, C., Schadt, E., and Ulrich, R. G. (2001) Clustering of hepatotoxins based on the

Fontaine et al. mechanism of toxicity using gene expression profiles. Toxicol. Appl. Pharmacol. 175, 28-42. (38) Gray, J. M., and Barnsley, E. A. (1971) The metabolism of crotyl phosphate, crotyl alcohol and crotonaldehyde. Xenobiotica 1, 55-67.

TX0255119